The gap in shared understanding of LLM capability is widening

本文为原文前 6,000 字符的节选翻译，完整内容请查看原文。

Andrej Karpathy@karpathyDusting off this tweet from April because this gap in shared understanding of LLM capability is widening. It’s now less “two groups of people speaking past each other” and more a sharp funnel.

Andrej Karpathy@karpathy 翻出四月份的这条推文，是因为人们对大语言模型（LLM）能力的认知差距正在不断扩大。现在的状况不再是“两群人鸡同鸭讲”，而更像是一个尖锐的漏斗。

Napkin math somewhere around 6B people (~75% of the population) have barely come in contact with LLMs at all. - Around 1-2B (~20%) are casual and infrequent users of free-tier ChatGPT-like products. This group sees derpy chatbots and treats them a bit like a better Google search, a writing aid, or etc. Maybe an agent tries to book you a flight. I have non-tech friends who (reasonably, imo) say they have not much use for it at all. Even many professionals outside of math&code are in this tier. For example, execs and many other functions spend a lot of their time talking to other people, so while they understand what is happening intellectually, it is still second-hand and a bit abstract.
粗略估算，大约有 60 亿人（约占总人口的 75%）几乎没有接触过大语言模型。大约有 10 到 20 亿人（约 20%）是免费版 ChatGPT 类产品的随意、低频用户。这群人看到的只是笨拙的聊天机器人，并把它们当作更好的谷歌搜索、写作辅助工具等来使用。也许某个智能体能帮你订张机票。我有一些非技术背景的朋友（我认为很合理地）表示，他们觉得这东西没什么大用处。即使是数学和编程领域之外的许多专业人士也处于这一层级。例如，高管和其他许多职能部门的人大部分时间都在与人交流，因此虽然他们在智力上理解正在发生的事情，但这仍然是二手信息，显得有些抽象。
Now we get to professional use of frontier-grade LLMs in math&code. Somewhere around 20M people (0.2%) see first-hand that large, complex projects that used to take them weeks/months can now be completed by agents with a prompt. Building apps, copying apps, translating apps, decompiling apps from binaries… This has all happened very quickly and recently - less than 1 year ago, I was writing code manually by hand, typing memorized computer code commands into a code editor character by character, occasionally pressing Tab to autocomplete a little chunk of code.
接下来是数学和编程领域对前沿级大语言模型的专业应用。大约有 2000 万人（0.2%）亲眼见证了过去需要数周或数月才能完成的大型复杂项目，现在只需一个提示词（prompt）就能由智能体完成。构建应用、复制应用、翻译应用、从二进制文件中反编译应用……这一切发生得非常迅速且近期——不到一年前，我还在手动编写代码，将背诵的计算机代码命令逐字输入代码编辑器，偶尔按一下 Tab 键来自动补全一小段代码。
And finally we get to the ~5,000 people (~0.00006%) with access to frontier-grade systems internally. The external world has seen the preview. It looks like swarms of thousands of agents collaborating over weeks on software mega projects: minting zero days, running cyber attacks and defenses at machine speeds, discovering new science, advancing the frontier of mathematics. Things that would have taken top professionals in the industry years of work. Meanwhile, human review and comprehension are starting to fall behind. For example, people are still involved in the “archeology” of the OpenAI-HF incident from many months ago. Mathematicians may be poring over the 722 manuscripts on frontier mathematics for a while.
最后是那大约 5000 人（约 0.00006%），他们可以在内部访问前沿级系统。外部世界看到的只是预览版。而内部看起来像是成千上万个智能体在数周内协作完成大型软件项目：挖掘零日漏洞、以机器速度进行网络攻防、发现新科学、推进数学前沿。这些工作如果由行业顶尖专业人士来做，需要数年时间。与此同时，人类的审查和理解能力开始落后。例如，人们至今仍在对几个月前的 OpenAI-HF 事件进行“考古”。数学家们可能还需要一段时间来钻研那 722 份关于前沿数学的手稿。

The funnel is driven by a combination of factors: 1. The impact scales with ambition, problem size, and horizon. A question with a paragraph answer barely stresses the system. You need a reservoir of big, difficult problems that you really care about. This aspect drives the consumer / professional dimension of the funnel. 2. The jaggedness of the system (which I have written about a lot separately). Capability peaks in domains that are digital, verifiable and economically valuable. This is because LLM capability emerges from reinforcement learning on verifiable rewards on a curated environment mixture driven by revenue potential. This aspect primarily drives the area (e.g. math&code) dimension of the funnel. 3. Access. Free-tier, paid-tier, internal. So this is the weirdness of the moment. The general public has mostly not interacted with these systems. When they have, it looks like a derpy chatbot. The majority of professionals still see only a modest uplift. And a small sliver of professionals are experiencing the vertigo of the curve going vertical. And it is all happening at the same time.

这个漏斗是由多种因素共同驱动的：1. 影响程度随雄心、问题规模和视野而扩展。一个只需一段话回答的问题几乎无法给系统带来压力。你需要一个你真正关心的、庞大且困难的问题库。这一方面驱动了漏斗的消费者/专业维度。2. 系统的参差不齐（我曾单独写过很多关于这一点的内容）。能力在数字化、可验证且具有经济价值的领域达到顶峰。这是因为大语言模型的能力源于在由收入潜力驱动的精选环境组合中，通过可验证奖励进行的强化学习。这一方面主要驱动了漏斗的领域（如数学和编程）维度。3. 访问权限。免费版、付费版、内部版。这就是当下的怪异之处。普通大众大多没有与这些系统互动过。当他们互动时，看到的只是一个笨拙的聊天机器人。大多数专业人士只看到了适度的提升。而一小部分专业人士正在经历曲线垂直上升带来的眩晕感。这一切都在同时发生。